I imagine they could, yes. Motivation to do so is another question entirely.
Considering myself more of an experienced amateur, and a fan of the Socratic Method, let me ask you a few questions:
What do you currently understand about the relationship between two atoms in a material and the mean square relative displacement
(i.e. sigma^2) associated with the distance between them?
What does it mean to have a mean-square-displacement? ..and then a mean-square-relative displacement?
How would correlated, anti-correlated and uncorrelated relative motion affect the msrd?
Now extend this thinking to an ensemble of atoms.
What might you expect to be the source of a msrd in a perfect crystal at finite temperature?
What additional contribution may result from imperfections?